© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 5378
    Prev
    Next
    INDEX
    Explanations

    phrases related to being physically damaged or in a state of disrepair

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "broken."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    atur
    -0.74
    gee
    -0.72
     appell
    -0.72
     Salman
    -0.71
     Heller
    -0.70
     designate
    -0.69
    ffer
    -0.67
    advertising
    -0.66
    respond
    -0.66
    iens
    -0.66
    POSITIVE LOGITS
     broken
    3.63
    broken
    2.28
     Broken
    2.10
     shattered
    2.05
     fractured
    1.95
     cracked
    1.75
     busted
    1.62
     breaking
    1.54
     smashed
    1.54
     broke
    1.51
    Activations Density 0.013%

    No Known Activations